PROJECT SUMMARY/ABSTRACT Alzheimer?s Disease (AD) is characterized by the presence of ?-amyloid plaques, tau-containing neurofibrillary tangles, and neuronal loss; yet, little is known about the development and spread of neurodegenerative processes. Recent progress in neuroimaging has enabled the detection of the main features of the AD pathophysiological cascade using positron emission tomography (PET) and magnetic resonance imaging (MRI), allowing researchers to track the development of the disease in vivo. Candidate & environment. The candidate?s career goal is to become an independent investigator and lead imaging research to elucidate the mechanisms underlying Alzheimer?s disease (AD) and develop precision medicine tools tailored towards individual patients. During this K award, the candidate will extend his expertise in multimodal neuroimaging by receiving additional training in connectivity analyses and will acquire expertise in network science and mathematical modeling. This new skill set will allow him to pursue innovative research projects testing hypotheses about the spread of AD pathology. The candidate will have access to large datasets of state-of-the-art PET and MRI data; he will be working with a multi-disciplinary team of world-renowned mentors and collaborators with expertise spanning imaging, behavioral neurology, bioengineering, pathology and statistics. The exceptional resources and diverse scientific community at the UCSF Memory and Aging Center will provide an ideal environment for the candidate?s training and will foster his growth as a future successful independent investigator. Research project. The research project proposed herein aims to use longitudinal human neuroimaging to test a disease model informed by basic science, postulating that i) tau originates in focal areas (?epicenters?) and progresses throughout the brain via pre-existing connections, and that ii) tau triggers local neuronal loss. Analyses will rely on PET with a radiotracer for tau pathology and structural MRI to measure brain atrophy, a proxy for neuronal loss, in patients in the early clinical stages of AD (mild cognitive impairment or dementia due to AD). The candidate will use an established network diffusion model to quantify how imaging abnormalities conform to and progress along the brain structural connectome. Specifically, he will use this diffusion model to infer prior disease states, i.e. identify tau epicenters in individual patients (Aim 1), and to predict future tau spread (Aim 2). Baseline tau-PET will then be used to predict future brain atrophy (Aim 3). The combined use of state- of-the-art multimodal imaging and mathematical modeling will not only allow the candidate to test hypotheses about mechanisms of disease spread, but will also inform the development of patient-tailored measures of disease progression with implications for the design of more efficient clinical trials.